1. Top of page
  2. Abstract
  3. References

See article in J. Gastroenterol. Hepatol. 2012; 27: 142–148.

Non-alcoholic fatty liver disease (NAFLD) consists of two clinical entities: simple steatosis, which generally follows a benign non-progressive clinical course, and non-alcoholic steatohepatitis (NASH), which could progress to cirrhosis and hepatocellular carcinoma. The differentiation of NASH from simple steatosis is clinically important, since the prognoses of these conditions clearly differ so diametrically. NASH has emerged as a clinicopathological entity, and even now biopsy evaluation is considered the “gold standard” for its definitive diagnosis.1

In 1999, Matteoni et al.2 described a classification system that served to distinguish non-NASH from NASH. They divided 132 NAFLD patients into four categories: type 1, steatosis alone; type 2, steatosis with lobular inflammation only; type 3, steatosis with hepatocellular ballooning; and type 4, type 3 plus either Mallory-Denk bodies or fibrosis. They confirmed the benign course of patients with type 1 or 2 NAFLD and the progressive course of patients who had either type 3 or 4 NAFLD. As a result of these differences, these authors defined types 1 and 2 histological forms of NAFLD as “non-NASH”, and types 3 and 4 as NASH. However, this classification did not include an assessment of the severity or pattern of NASH, such as the degree of inflammation, its location (e.g. lobular or portal), and the degree of fibrosis or its zonality.

In the same year as the Matteoni classification system was published, Elizabeth Brunt et al.3 proposed a grading and staging system for the semiquantitative evaluation of NASH. She followed the same paradigm as used in chronic hepatitis for separately grading activity, and staging fibrosis. It needs to be emphasized that this original “Brunt system” applied only to NASH; it was not applicable to the entire spectrum of NAFLD.

In 2005, the NASH Clinical Research Network (NASH-CRN) Pathology Committee developed and validated a histological feature scoring system, a NAFLD activity score (NAS) that addresses the full spectrum of NAFLD.4 As discussed later, there was a particular utilitarian purpose for developing this system, namely to have a semiquantitative instrument by which to judge treatment responses or disease progression in clinical studies. However, the score was developed to be applicable to adult and pediatric NAFLD patients, between whom some different patterns are observed. The score (NAS) is defined as the unweighted sum of the scores for steatosis (0–3), lobular inflammation (0–3), and ballooning (0–2). A score of > 5 correlated with a diagnosis of NASH made independently by an experienced pathologist without using the score; likewise, scores of less than 3 correlated with “not NASH,” and scores of 3 or 4 were regarded as borderline. Regarding fibrosis, stage 1 was perisinusoidal fibrosis in zone 3 (perivenular area; delicate [1A] and dense [1B]), and detection of portal fibrosis without perisinusoidal fibrosis was 1C. Stage 2 was characterized by additional evidence of focal or extensive periportal fibrosis. Stage 3 was bridging fibrosis, and stage 4 was cirrhosis. Some pathologists mistakenly use other scoring systems for fibrosis staging in NASH, but it is acknowledged that stages 2, 3, and 4 will be roughly equivalent to Metavir F2, F3, and F4.

Recently, two important articles on histological classifications for NAFLD have been published by those involved in this field who contributed to the two earlier reports.5,6 First, Brunt and colleagues have written a thoughtful review, entitled “NAS and the histopathological diagnosis in NAFLD: Distinct clinicopathological meanings”.5 Although the authors reminded us that the NAS system was never intended to be used for the diagnosis of NASH, NAS has frequently been used as a surrogate for establishing a diagnosis of NASH. Then, to evaluate whether this unintended use of the NAS is valid, biopsy and clinical data from the 976 adults in NASH-CRN studies were reviewed. They concluded that the definitive diagnosis of steatohepatitis (in this case, NASH) does not always correlate with threshold values of the NAS. Clinical trials and observational studies should take these different performance characteristics into account, and clinical pathologists should be encouraged to not use NAS as a categorical approach to diagnosis that has not considered the global impact of liver pathology. It is acknowledged that inexperienced liver pathologists might want to still measure the NAS, and perhaps review and discuss individual biopsies with more experienced colleagues when there appears to be substantive differences between the global impression of NASH versus not NASH and the actual NAS score.

The second important article, by Younossi et al.6 is entitled “Pathologic criteria for NASH: interprotocol agreement and ability to predict liver-related mortality”. The aims of this study were: (i) to assess the agreement of four pathological criteria for NASH, including Matteoni's criteria, NAS, Brunt's criteria, and the current study's criteria; and (ii) to determine their ability to predict long-term liver-related mortality. The study cohort consisted of 209 patients with biopsy-proven NAFLD who had at least 5 years of follow up. For this article, Brunt et al7 published a comment, entitled “Misuse of scoring systems”. They emphasize that the Brunt and NAS systems, as applied in this article, have not been appropriately used.

We are particularly interested in the latter part of the Younossi study. Although a number of pathological features were associated with liver-related mortality in the univariate analysis, fibrosis remained the only independent predictor of liver-related mortality on the multivariate analysis. According to this study, the presence and severity of fibrosis, rather than the diagnosis of NASH, dictate liver-related mortality. This issue is very important, since we believe that “NASH is the progressive form of NAFLD”, and clinically, liver-related mortality is the most important outcome. Angulo8 emphasized the importance of liver fibrosis in the Editorial for this article. Histological fibrosis stages are easy to diagnose with high concordance among pathologists. It should now be considered why NAS is so frequently used as pathological criteria for NASH.

In this issue of the Journal of Gastroenterology and Hepatology,9 Kyung et al. also used NAS for diagnostic criteria. It seems to us that clear, easy, and reproducible standard criteria for the histological classification of NASH are still required, not only for expert liver pathologists, but also for the general pathologists throughout the world who will be most involved with assisting clinicians in diagnostic classification of patients with fatty liver diseases.

Another important histological injury in NASH is portal-based lesions. It has been reported that many pediatric cases have atypical features, such as more periportal steatosis (zone 1), little or no ballooning or Mallory hyaline, and more portal-based chronic inflammation and fibrosis. In contrast, a predominantly zone 3 injury pattern is typically seen in adult patients with NASH. Schwimmer et al.10 demonstrated two different forms of steatohepatitis: type 1 adult-like NASH type was characterized by steatosis, ballooning degeneration and perisinusoidal fibrosis, while type 2 (zone 1 predominance or panacinar type) was characterized by steatosis, portal inflammation, and portal fibrosis. The characteristic features of type 2 NASH were as follows: more frequent among boys, younger children, and severely obese children, and association with more severe fibrosis. Type 1 NASH was more common in white children, whereas type 2 NASH was more common in children of Asian, Native American, and Hispanic ethnicity. There were no differences in laboratory data between patients with types 1 and 2 NASH. A similar pattern emerged among children enrolled in the NASH-CRN11.

Table 1 shows the prevalence of simple steatosis, NASH types 1 and 2, and overlap. In Schwimmer's study,10 type 2 was the most common pattern. In contrast, type 2 was not the most common type, and type 1 was rarely seen in Nobili's study12 of Italian children aged 9–18 years. In the CRN study11 and Ko's study13 from Korea, overlap was not assessed. The study from Ko13 showed that type 2 was the most common form and was associated with greater severity of obesity and advanced fibrosis. In Kyung's study in Korean young men,9 more than half of cases of NASH exhibited overlap type, and fibrosis was prominent in this type with high NAS score. The ages of type 2 patients were younger than those of overlap type. The authors speculated that overlap can progress from pediatric-type NASH; the pattern in the middle of transition from pediatric to adult type. The authors first performed a cross-sectional study among young men and studied many aspects of clinicopathological features. Therefore their findings are very important. They elucidate the role of overlap type. To confirm this hypothesis, understanding of the natural history of pediatric cases of NAFLD is needed.

Table 1.  Prevalence of simple steatosis, non-alcoholic steatohepatitis types 1 and 2, and overlap by Schwimmer's classification
 Simple steatosisType 1 adult likeType 2 portal predominanceOverlap
Mean ageMale sex (%)(%)Mean ageMale sex (%)(%)Mean ageMale sex (%)(%)Mean ageMale sex (%)(%)
  1. ND, no data.

Schwimmer et al.1012.362.51613.5351711.5755111.56916
n = 100
Patton et al.1113.16724.213.1784311.18632.9NDNDND
n = 149
Nobili et al.12N.D.ND16.7NDND2.4NDND28.6NDND52.4
n = 84
Ko et al.1310.488.922.512.681.533.812.382.943.8NDNDN.D.
n = 80
Kyung et al.922.4ND7.821.8ND26.620.8ND6.322.4ND59.4
n = 64

Brunt et al.14 revealed that that the majority of biopsies with NAFLD have at least some portal inflammation, and that more than mild portal inflammation correlates with advanced fibrosis. It has also been reported that ductular reaction is associated strongly with fibrosis.15 Previously perivenular changes were highlighted in NASH. These findings show that zone 1 changes are very interesting and might be important in NASH pathogenesis. In addition, as the lipotoxic concept of NASH gains strength, and genetic predisposition is being better characterized, correlations between such etiopathogenic determinants of NAFLD and disease pathological phenotype will be of considerable interest.


  1. Top of page
  2. Abstract
  3. References